We start our student contributed guest post series with a look at data from a topological lens. Skip Moses graduated from Chico State Department of Mathematics and Statistics in 2022. He triple majored in Pure Math, Applied Math, Statistics, and completed the Certificate in Data Science.
Data often has an underlying structure or geometry that can be modeled as a signal on the vertices of a weighted, undirected graph. Dong et. al. provide an algorithm for learning the underlying structure of a graph given a smooth signal representation on the graph here. Dong et. al. originally implemented the algorithm in MATLAB, however, it can be solved effiecently in Python. In the post here a brief description of the algorithm, and example of a learned graph are provided.